Ok. So that is the general idea and with examples taken from Mazak Vertical Center ( SMART).
first of all we need this Alarm 55 IO with the number that is attached to it, as it says what station and system are "suspected" to be faulty. That is the funny part of it. Your machine have systems and stations for Remote IO boards. Once system is configured, stations are connected in a chain. If one station is troublemaker ( say, somewhere in the middle of the chain) than all other station are not visible to NC. Sometimes you can get error pointing to one particular station , while in reality the faulty one is connected before this one.
Here are pictures form Smart mill where you can see what systems are and stations as well. Also a screenshot of DIAGNOSTIC monitor to check status of diagnostic registers R5xxx. You need to do the same and see if you have any numbers growing really really big.
You can see the page where all diagnostic registers are listed and also this info:
SIGNAL NAME:
-max value of communication ( probably , how many times NC lost communication with this particular IO board and have managed to reestablish it later)
-Max value CRC ( that is redundancy code to check if any data was lost during communication phase )
Accumulated values are the same but since the machine creating.
If you see any numbers mad big or whatever, you can probably swap remote IO boards ( don't forget to change rotary switch address) and see if problem moves.
if it moves , you will see the same error 55 IO remote , but with different number attached to it, pointing to different location ( physical location ).
That is the only way to troubleshoot the problem, unless you have someone from Mitsubishi and he has some other " classified " information, not available to other mortals.
P.S in my example, we have system 1 and 3 with 3 stations attached. And as you can see, R5xxx registers correspond to my configurations. So I assume, that this applicable across all Mazaks.